INDEX

Explanations

mentions of shapes or words related to shaping

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "shape" 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

artment

-0.72

amily

-0.71

govtrack

-0.71

 Edited

-0.68

unts

-0.68

GS

-0.65

Mub

-0.65

BILITIES

-0.65

nea

-0.64

uers

-0.63

POSITIVE LOGITS

shape

1.04

 shape

0.96

 shif

0.88

cut

0.87

 shapes

0.86

 Shape

0.82

Shape

0.79

 shaped

0.77

sheet

0.75

lier

0.73

Activations Density 0.021%

mentions of shapes or words related to shaping

mentions of the word "shape" in various contexts

No Comments

No Known Activations