INDEX

Explanations

mentions of the word "Fury" or related terms

oai_token-act-pair · gpt-3.5-turbo

references to anger or strong emotional expressions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ership

-0.86

organic

-0.84

opathy

-0.80

icals

-0.80

pocket

-0.77

izational

-0.77

esthetic

-0.76

nai

-0.72

kai

-0.72

repeat

-0.72

POSITIVE LOGITS

 Fury

1.13

 tremend

0.82

anche

0.81

ricanes

0.76

 Packs

0.75

FC

0.75

ous

0.72

 Beasts

0.71

mares

0.68

oshenko

0.68

Activations Density 0.008%

mentions of the word "Fury" or related terms

references to anger or strong emotional expressions

No Comments

No Known Activations