INDEX

Explanations

words related to negative impact or harm

oai_token-act-pair · gpt-3.5-turbo

expressions related to causing harm or damage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

mad

-0.72

uls

-0.71

dding

-0.67

uesday

-0.66

ãĥ¼ãĥ³

-0.66

igers

-0.66

ricks

-0.65

ellen

-0.65

odor

-0.63

leans

-0.63

POSITIVE LOGITS

 havoc

1.11

 credibility

1.02

 delicate

0.96

 morale

0.94

 livelihood

0.90

 sensibilities

0.90

 friendships

0.88

 integrity

0.88

 innocent

0.87

 morals

0.85

Activations Density 0.229%

words related to negative impact or harm

expressions related to causing harm or damage

No Comments

No Known Activations