INDEX

Explanations

words related to spatial concepts, specifically indicating a location or direction

oai_token-act-pair · gpt-3.5-turbo

references to "the other side" 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

GM

-0.67

Deal

-0.66

 Entered

-0.65

ACP

-0.61

 Inher

-0.60

Coach

-0.59

 Playoff

-0.57

 Serve

-0.57

essed

-0.56

Grab

-0.56

POSITIVE LOGITS

worldly

0.78

than

0.73

 altogether

0.69

ottest

0.66

arie

0.65

 than

0.64

blance

0.62

bang

0.61

arettes

0.60

axis

0.60

Activations Density 0.107%

words related to spatial concepts, specifically indicating a location or direction

references to "the other side" in various contexts

No Comments

No Known Activations