INDEX

Explanations

words related to relationships and belonging

oai_token-act-pair · gpt-3.5-turbo

the repeated use of the word "that" 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 @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ummer

-0.69

onsequ

-0.69

ising

-0.66

uty

-0.65

tnc

-0.65

detail

-0.65

basic

-0.63

events

-0.61

EngineDebug

-0.61

ittal

-0.59

POSITIVE LOGITS

 specialize

1.00

 owns

0.97

 participated

0.93

 specializes

0.90

 cares

0.89

 loves

0.85

 reside

0.85

 inhabit

0.85

 wants

0.84

 cared

0.82

Activations Density 0.143%

words related to relationships and belonging

the repeated use of the word "that" in various contexts

No Comments

No Known Activations