INDEX

Explanations

words related to locations or proper nouns

oai_token-act-pair · gpt-3.5-turbo

periods or some form of punctuation

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 withd

-0.87

 cabbage

-0.82

 inver

-0.82

 nesting

-0.79

 flowering

-0.78

awa

-0.78

 challeng

-0.78

 attacker

-0.76

 tack

-0.74

 crabs

-0.74

POSITIVE LOGITS

 Accessed

1.19

jpg

1.07

 Retrieved

1.04

Org

1.00

txt

0.96

org

0.94

 Located

0.93

esp

0.92

 Learns

0.92

xxx

0.91

Activations Density 0.360%

words related to locations or proper nouns

periods or some form of punctuation

No Comments

No Known Activations