INDEX

Explanations

phrases related to reasons or causes

oai_token-act-pair · gpt-3.5-turbo

phrases indicating partial causes or reasons for events or conditions

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

eer

-0.94

Trend

-0.82

eers

-0.80

 clipboard

-0.75

ãĥ¤

-0.74

ciating

-0.73

verbs

-0.72

endi

-0.72

ERY

-0.70

STER

-0.69

POSITIVE LOGITS

 obscured

0.88

 cloudy

0.85

 overlapping

0.81

 veiled

0.71

due

0.70

 blinded

0.70

 reflecting

0.67

 opaque

0.66

 compensate

0.65

heartedly

0.64

Activations Density 0.011%

phrases related to reasons or causes

phrases indicating partial causes or reasons for events or conditions

No Comments

No Known Activations