INDEX

Explanations

instances of the word "Choosing" or variations of it

oai_token-act-pair · gpt-3.5-turbo

the presence of the word "Cho" and its variations in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 inaccur

-0.62

 tremend

-0.62

 Aram

-0.60

 seeker

-0.59

sburg

-0.58

TERN

-0.58

self

-0.56

Ast

-0.55

venture

-0.54

master

-0.54

POSITIVE LOGITS

osing

1.56

ices

1.44

pper

1.35

pped

1.35

ppy

1.31

pping

1.29

oser

1.28

oses

1.20

ppers

1.20

ose

1.10

Activations Density 0.027%

instances of the word "Choosing" or variations of it

the presence of the word "Cho" and its variations in various contexts

No Comments

No Known Activations