INDEX

Explanations

terms related to academic context or academia

oai_token-act-pair · gpt-3.5-turbo

terms related to academic performance and entitlement

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

 Bundy

-0.78

lessly

-0.73

doors

-0.73

lda

-0.72

ACTED

-0.72

generic

-0.70

xes

-0.70

xon

-0.70

coni

-0.69

ality

-0.66

POSITIVE LOGITS

ulty

0.96

VERTISEMENT

0.91

ademic

0.82

iott

0.82

 attainment

0.82

 institutions

0.78

 academic

0.77

 literature

0.76

 disciplines

0.76

 misconduct

0.75

Activations Density 0.012%

terms related to academic context or academia

terms related to academic performance and entitlement

No Comments

No Known Activations