INDEX

Explanations

action verbs related to investigating, analyzing, and studying

oai_token-act-pair · gpt-3.5-turbo

research-related terminology associated with investigation and analysis

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

Lago

-0.73

jar

-0.68

Jar

-0.66

usercontent

-0.65

Notice

-0.63

Allah

-0.63

mint

-0.63

scribe

-0.62

ridden

-0.62

argon

-0.62

POSITIVE LOGITS

 hypotheses

0.86

 feasibility

0.80

 statistical

0.77

 correlations

0.76

 anomalies

0.74

 findings

0.73

 empir

0.72

 probing

0.72

 trends

0.72

 quant

0.69

Activations Density 0.156%

action verbs related to investigating, analyzing, and studying

research-related terminology associated with investigation and analysis

No Comments

No Known Activations