INDEX

Explanations

terms related to admissions, such as admissions to schools, hospitals, events, or programs

oai_token-act-pair · gpt-3.5-turbo

references to admissions in various contexts

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

rior

-0.79

hed

-0.71

enegger

-0.70

 limb

-0.67

ector

-0.67

aptic

-0.66

VALUE

-0.65

 nesting

-0.64

lette

-0.64

estamp

-0.63

POSITIVE LOGITS

 Admission

0.99

aneers

0.94

 admission

0.88

ysis

0.77

prise

0.76

gow

0.76

ibility

0.76

ulty

0.74

glers

0.74

Reviewer

0.72

Activations Density 0.026%

terms related to admissions, such as admissions to schools, hospitals, events, or programs

references to admissions in various contexts

No Comments

No Known Activations