INDEX

Explanations

references to different weights or weight-related terms

oai_token-act-pair · gpt-3.5-turbo

references to weight 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 @ 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

ICLE

-0.84

BLE

-0.82

VIS

-0.73

 Noir

-0.73

Mb

-0.73

Suc

-0.72

xon

-0.70

 Kidd

-0.69

UTE

-0.69

 Gutenberg

-0.68

POSITIVE LOGITS

 weights

1.05

lifting

1.03

weight

0.98

 weight

0.96

 heaviest

0.93

weights

0.92

 heavier

0.89

 burdens

0.89

iless

0.87

 lifting

0.86

Activations Density 0.021%

references to different weights or weight-related terms

references to weight in various contexts

No Comments

No Known Activations