INDEX

Explanations

phrases indicating further information or advancement

oai_token-act-pair · gpt-3.5-turbo

instances of the word "further" and its variations, indicating a focus on continuation or additional information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ldom

-0.81

cius

-0.76

pei

-0.75

ertodd

-0.72

EA

-0.71

orno

-0.69

pas

-0.69

odor

-0.68

enos

-0.68

liest

-0.68

POSITIVE LOGITS

ado

0.96

ance

0.94

 Reading

0.88

 refinement

0.87

 compl

0.85

 subdiv

0.84

 developments

0.79

 investigation

0.78

 refining

0.76

 deterioration

0.76

Activations Density 0.049%

phrases indicating further information or advancement

instances of the word "further" and its variations, indicating a focus on continuation or additional information

No Comments

No Known Activations