INDEX

Explanations

words related to physical locations or geographic entities

oai_token-act-pair · gpt-3.5-turbo

proper nouns and names of individuals or entities

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

 [âĢ¦]

-0.60

 multim

-0.59

***

-0.59

 [...]

-0.58

 curve

-0.56

apt

-0.55

 class

-0.55

 smack

-0.55

[/

-0.54

 incoming

-0.54

POSITIVE LOGITS

ÃŃn

0.86

orst

0.86

orah

0.80

kov

0.78

itan

0.77

uter

0.77

ruary

0.76

onso

0.75

yon

0.75

Ã«

0.73

Activations Density 0.417%

words related to physical locations or geographic entities

proper nouns and names of individuals or entities

No Comments

No Known Activations