INDEX

Explanations

statements expressing differing levels of correctness or morality, along with recommendations or judgments about actions

oai_token-act-pair · gpt-3.5-turbo

phrases that express mistakes or wrongdoings related to moral or ethical judgments

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 strengths

-0.70

lator

-0.68

 reperto

-0.67

aukee

-0.63

 calmed

-0.63

 linem

-0.63

 rapport

-0.62

hani

-0.62

uli

-0.61

 delightful

-0.61

POSITIVE LOGITS

 underestimate

0.88

 knowingly

0.83

 anymore

0.81

 presume

0.80

 oppose

0.79

 condone

0.78

 whatsoever

0.77

 anyone

0.75

 accuse

0.75

 impose

0.75

Activations Density 0.207%

statements expressing differing levels of correctness or morality, along with recommendations or judgments about actions

phrases that express mistakes or wrongdoings related to moral or ethical judgments

No Comments

No Known Activations