INDEX

Explanations

terms related to opinions and organizational entities like unions

oai_token-act-pair · gpt-3.5-turbo

variations of the word "opinion" and related terms

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

weights

-0.70

 floor

-0.66

 fabrication

-0.64

ten

-0.63

abases

-0.61

\\\\\\\\\\\\\\\\

-0.59

cases

-0.59

ropolis

-0.58

WT

-0.57

 Tanz

-0.56

POSITIVE LOGITS

hip

1.00

piracy

0.99

pread

0.98

hips

0.98

ervatives

0.97

naire

0.94

hov

0.94

ervative

0.93

ystem

0.92

chool

0.86

Activations Density 0.050%

terms related to opinions and organizational entities like unions

variations of the word "opinion" and related terms

No Comments

No Known Activations