INDEX

Explanations

high-level concepts related to comparison or evaluation

oai_token-act-pair · gpt-3.5-turbo

phrases that reference comparisons or evaluations in relation to a particular context

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

resent

-0.72

avorite

-0.69

yden

-0.67

****************

-0.64

oute

-0.63

dinand

-0.62

 Bene

-0.62

 tatt

-0.61

 Rothschild

-0.61

ried

-0.60

POSITIVE LOGITS

pring

0.90

pace

0.88

ames

0.84

eme

0.84

uman

0.83

peed

0.79

cale

0.78

cape

0.76

 terms

0.73

paced

0.72

Activations Density 0.022%

high-level concepts related to comparison or evaluation

phrases that reference comparisons or evaluations in relation to a particular context

No Comments

No Known Activations