INDEX

Explanations

specific details, references, or citations within text

oai_token-act-pair · gpt-3.5-turbo

references to the act of citing or quoting sources

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

soever

-0.69

adia

-0.67

ados

-0.63

apon

-0.61

 Canad

-0.61

 Madden

-0.60

deg

-0.60

 Brenda

-0.60

think

-0.58

mAh

-0.57

POSITIVE LOGITS

 quotation

0.85

 wording

0.82

 verse

0.80

zinski

0.79

 specifics

0.79

 anecdote

0.76

 specificity

0.76

inventoryQuantity

0.75

 phrase

0.74

 quotes

0.74

Activations Density 0.264%

specific details, references, or citations within text

references to the act of citing or quoting sources

No Comments

No Known Activations