INDEX

Explanations

names related to Russian locations or people

oai_token-act-pair · gpt-3.5-turbo

occurrences of specific suffixes and prefixes in words

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

footed

-0.71

vae

-0.70

DOI

-0.65

avorite

-0.61

)].

-0.60

SPONSORED

-0.59

 confounding

-0.57

 Jinn

-0.57

 hemor

-0.57

 undermin

-0.57

POSITIVE LOGITS

roth

0.92

opol

0.77

enne

0.75

ral

0.74

»Ĵ

0.69

lic

0.69

arte

0.66

ene

0.66

ela

0.65

ijn

0.65

Activations Density 0.089%

names related to Russian locations or people

occurrences of specific suffixes and prefixes in words

No Comments

No Known Activations