INDEX

Explanations

instances where a comparison or decision between multiple options is being made

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "which."

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

GROUND

-0.81

Balt

-0.73

 Glob

-0.72

Rog

-0.71

rise

-0.71

Adams

-0.71

athi

-0.70

rene

-0.69

ben

-0.66

inence

-0.64

POSITIVE LOGITS

 kinds

0.84

 sorts

0.82

 contingency

0.75

soever

0.71

iple

0.70

 flavors

0.69

 delim

0.68

 types

0.67

 redes

0.66

 species

0.64

Activations Density 0.052%

instances where a comparison or decision between multiple options is being made

occurrences of the word "which."

No Comments

No Known Activations