INDEX

Explanations

references to large groups or entities

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "large" in various contexts

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

qi

-0.80

pered

-0.78

gnu

-0.72

keeper

-0.69

Reviewed

-0.67

alth

-0.66

per

-0.65

CB

-0.64

tein

-0.64

pering

-0.64

POSITIVE LOGITS

 intestine

0.89

 proport

0.78

 scale

0.76

 denomination

0.75

(>

0.75

 quantities

0.74

 gatherings

0.74

 amounts

0.74

 eleph

0.70

pox

0.69

Activations Density 0.011%

references to large groups or entities

occurrences of the word "large" in various contexts

No Comments

No Known Activations