INDEX

Explanations

sentences where the concept of 'how' is being addressed or discussed

oai_token-act-pair · gpt-3.5-turbo

inquiries related to the concept of "how."

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

agonists

-0.75

ceptions

-0.72

odder

-0.68

 ........

-0.65

ception

-0.64

 Tanz

-0.62

UME

-0.61

wear

-0.61

agonist

-0.60

aughed

-0.60

POSITIVE LOGITS

soever

0.82

beit

0.74

HCR

0.73

ells

0.73

 much

0.73

ever

0.71

 exactly

0.70

 situational

0.68

ls

0.67

 badly

0.66

Activations Density 0.080%

sentences where the concept of 'how' is being addressed or discussed

inquiries related to the concept of "how."

No Comments

No Known Activations