INDEX

Explanations

proper nouns, specifically related to sports teams, technology, and locations

oai_token-act-pair · gpt-3.5-turbo

references to sports teams and events

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

accompanied

-0.76

 comprom

-0.67

uggest

-0.62

 tatt

-0.61

 fingert

-0.60

 ÂŃ

-0.57

 caution

-0.57

 speeding

-0.57

 rave

-0.57

 alike

-0.56

POSITIVE LOGITS

.............

0.74

0.73

è¦ļéĨĴ

0.72

Shot

0.72

 Blvd

0.72

Rd

0.69

 ..............

0.68

lvl

0.68

 ----

0.66

 âĵĺ

0.66

Activations Density 0.418%

proper nouns, specifically related to sports teams, technology, and locations

references to sports teams and events

No Comments

No Known Activations