INDEX

Explanations

dates in specific formats

oai_token-act-pair · gpt-3.5-turbo

ordinal numbers or date indicators

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 streng

-0.68

ioned

-0.62

Marginal

-0.61

 tranquil

-0.60

Transfer

-0.60

gypt

-0.59

ufact

-0.59

BIL

-0.57

una

-0.56

 counselling

-0.56

POSITIVE LOGITS

1.23

1.21

1.14

1.13

1.12

1.11

1.06

1.04

1.03

Activations Density 0.026%

dates in specific formats

ordinal numbers or date indicators

No Comments

No Known Activations