INDEX

Explanations

phrases indicating clear conclusions or evaluations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "clearly" emphasizing transparency or obviousness in statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

uese

-0.79

aily

-0.76

oleon

-0.70

anish

-0.68

umption

-0.68

hell

-0.68

awaru

-0.68

lav

-0.67

urch

-0.67

rost

-0.67

POSITIVE LOGITS

 deline

0.97

 marked

0.84

 identifiable

0.82

 distinguish

0.80

 differentiated

0.78

 readable

0.75

 articulated

0.74

 outwe

0.74

 differentiate

0.74

 spelled

0.73

Activations Density 0.027%

phrases indicating clear conclusions or evaluations

instances of the word "clearly" emphasizing transparency or obviousness in statements

No Comments

No Known Activations