INDEX

Explanations

words containing the subsequence "rep"

oai_token-act-pair · gpt-3.5-turbo

instances of the term "intrepid" and related forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Reborn

-0.73

 Seym

-0.66

 generously

-0.62

steen

-0.60

 Panda

-0.60

 Cosmos

-0.60

ãĢĲ

-0.60

 Mermaid

-0.59

 Reincarn

-0.58

 flats

-0.58

POSITIVE LOGITS

LECT

0.82

aneous

0.80

ACTION

0.80

lycer

0.79

elta

0.76

igent

0.73

ATER

0.69

igil

0.66

ention

0.66

thora

0.66

Activations Density 0.049%

words containing the subsequence "rep"

instances of the term "intrepid" and related forms

No Comments

No Known Activations