INDEX

Explanations

names starting with 'Se'

oai_token-act-pair · gpt-3.5-turbo

specific names or identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

:,

-0.80

relative

-0.77

antine

-0.76

nil

-0.71

adic

-0.70

 covari

-0.67

',

-0.65

less

-0.64

 USSR

-0.64

 solitary

-0.63

POSITIVE LOGITS

rez

0.80

 Eater

0.78

sale

0.69

reon

0.66

eno

0.65

regon

0.65

oru

0.65

omach

0.65

ophe

0.65

illard

0.63

Activations Density 0.000%

names starting with 'Se'

specific names or identifiers

No Comments

No Known Activations