INDEX

Explanations

words related to possibility or suggestion

oai_token-act-pair · gpt-3.5-turbo

phrases that include the word "well"

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

noticed

-0.67

ament

-0.60

quartered

-0.60

cutting

-0.59

 Canary

-0.59

oshenko

-0.59

 Cutting

-0.58

processing

-0.57

ressing

-0.57

assembly

-0.57

POSITIVE LOGITS

be

1.10

owe

0.91

 earn

0.85

 derive

0.85

 prove

0.84

 suffice

0.83

 resemble

0.82

BE

0.80

 have

0.80

 Marketable

0.79

Activations Density 0.052%

words related to possibility or suggestion

phrases that include the word "well"

No Comments

No Known Activations