INDEX

Explanations

numerical ranges, particularly related to quantities of people or groups

oai_token-act-pair · gpt-3.5-turbo

terms related to group sizes or team compositions

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

reck

-0.65

cross

-0.65

entimes

-0.63

urus

-0.62

ename

-0.62

hend

-0.62

 Shank

-0.61

ends

-0.61

 Compare

-0.60

SPONSORED

-0.60

POSITIVE LOGITS

enger

0.76

acters

0.75

clusive

0.74

atan

0.74

holder

0.71

doms

0.69

owitz

0.69

volent

0.69

essee

0.66

 delegation

0.66

Activations Density 0.033%

numerical ranges, particularly related to quantities of people or groups

terms related to group sizes or team compositions

No Comments

No Known Activations