INDEX

Explanations

instances where the word "Sentinel" is mentioned

oai_token-act-pair · gpt-3.5-turbo

references to the "Sentinel" entity or project

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

sen

-0.77

roma

-0.75

ulates

-0.74

do

-0.73

mers

-0.73

arded

-0.72

orne

-0.71

romy

-0.70

si

-0.69

rowing

-0.69

POSITIVE LOGITS

 Sentinel

1.23

inel

0.93

 CrossRef

0.81

 Extensions

0.76

 Superior

0.76

andum

0.75

 Hers

0.71

 Advocate

0.71

 Protective

0.70

 newsp

0.67

Activations Density 0.015%

instances where the word "Sentinel" is mentioned

references to the "Sentinel" entity or project

No Comments

No Known Activations