INDEX

Explanations

percentages or numerical changes

oai_token-act-pair · gpt-3.5-turbo

numerical ranges and comparisons

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Exit

-0.77

luck

-0.68

CI

-0.68

wcsstore

-0.67

Marginal

-0.67

Upgrade

-0.66

Allow

-0.64

Delivery

-0.64

WER

-0.63

WIND

-0.62

POSITIVE LOGITS

 something

0.77

 fewer

0.71

 nearly

0.70

 merely

0.70

 mere

0.68

 roughly

0.65

 almost

0.65

 just

0.64

 simply

0.64

 barely

0.63

Activations Density 0.053%

percentages or numerical changes

numerical ranges and comparisons

No Comments

No Known Activations